Deriving star formation histories from photometric spectral energy distributions with diffusion k-means
Abstract
The star formation histories of galaxies give us insight into how galaxies have changed over time and continue to evolve as factories of the universe's gas, dust, and metal content. We can measure these star formation histories coarsely from integrated spectra of galaxies, but as we look back farther in the universe's history even marginal signal to noise spectra become costly. We have shown in previous works that a reduced basis set of averaged stellar populations determined by the machine learning diffusion k-means (DFK) algorithm can be used to recover precise and accurate star formation histories from low signal to noise spectra. In this work, we use this reduced basis set to analyze photometric galaxy spectral energy distributions (SEDs) that may be available in the absence of a spectrum. To compare a DFK basis set to current methods used to analyze galaxy SEDs, we look at the 3D-HST photometry catalog of galaxies observed in the GOODS-N field. We compare the stellar population results using FAST from the catalog to the results using the DFK basis set. Precise and accurate stellar populations from photometric SEDs using the DFK basis set would provide a unique tool for analyzing galaxy star formations histories out to high redshift.
- Publication:
-
American Astronomical Society Meeting Abstracts #233
- Pub Date:
- January 2019
- Bibcode:
- 2019AAS...23314429M